Study concludes that NSF is widespread in high-risk patients, but diagnostic method raises questions

Contrary to conventional wisdom, a new study suggests the small number of NSF cases known today may be just the tip of the iceberg. In fact, the disease appears to be prevalent - of "epidemic" proportions - in hemodialysis patients, according to rheumatologists at Massachusetts General Hospital.

In the study of dialysis patients attending six outpatient centers, 13% overall and 30% known to have undergone contrast-enhanced MRI tested positive for NSF on skin tests, according to the results, published in Arthritis & Rheumatism, the official journal of the American College of Rheumatology, in October (Study finds NSF more common than previously reported).

The study found that mortality rates were three times higher in the NSF patients, a striking finding that caught the attention of Reuters and U.S. News & World Report, among others in the mainstream media.

In light of the risks, Mass General does not generally administer any kind of gadolinium in hemodialysis patients, said Dr. Jonathan Kay, one of the study's authors. In the rare cases where it is deemed necessary, providers obtain informed consent.

An editorial accompanying the article, by experts at Yale University, suggested the higher incidence could reflect detection of a less severe form of NSF. But according to Kay, clinical director of MGH's rheumatology unit, the mortality rates indicate otherwise.

"We were struck by the prevalence of these skin changes. We were most surprised by the marked increase in mortality, which was certainly consistent with severe NSF in many of the patients," said Kay in an interview with Diagnostic Imaging.

The words "epidemic" and "prevalent" are not usually associated with NSF. In the disease's short history, it has more typically been labeled "extremely rare." Recently, reports of new NSF cases have been dwindling and awareness among clinicians has been running high, according to experts.

In a Sept. 12 letter to health professionals, the four manufacturers of five U.S.-approved contrast agents noted that just 250 cases of gadolinium-related NSF had been reported to the FDA. An estimated 135 million doses of gadolinium have been administered worldwide, according to industry sources.

The letter also cited a study showing incidence of 4% in patients with renal insufficiency following gadolinium exposure and restated terms of a black box warning on all gadolinium agents announced in May (The NSF 'black box' holds danger and opportunity).

Bayer Schering, the manufacturer of the agent used in the Mass General study (Magnevist), points out that the FDA has advised that biopsy is necessary to confirm a diagnosis, though biopsy does not appear to be required for reporting adverse events through the Medwatch system.

"Currently, the only clinically accepted way to distinguish NSF from other skin conditions is through the identification of features that are unique to NSF via histopathology (on a deep skin biopsy specimen) rather than on clinical skin findings alone, which may also be present in fibrosing skin conditions other than NSF," said Marcy Funk, company spokesperson.

It's difficult to compare the rates determined via skin biopsy with reports based on the novel approach of evaluating cutaneous skin changes, Funk said.

However, biopsies pose a risk to some patients on hemodialysis and using skin tests instead boosted study participation rates to very high levels, the authors reported. Furthermore, Kay told Diagnostic Imaging, the skin testing process used in the trial to make the diagnosis is accepted in clinical practice.

Dr. Emmanuel Kanal, head of the American College of Radiology Blue Ribbon Panel on MR Safety said it is quite likely that the incidence of NSF is higher than presently detected for several reasons. These include the possibility of subclinical cases of NSF, relative lack of awareness of the entity and how to diagnose it, and general underreporting of cases to a central database or authority.

"This manuscript provides evidence that, utilizing skin lesion characterization alone, NSF may be more prevalent than had been initially suspected," Kanal said. "Of course, while the simplicity of the extremity skin characterization of hyperpigmentation, skin hardening, and tethering - where any two of three is considered positive - is quite inviting, this simplicity itself needs to be more rigorously verified with skin biopsy."

Kanal noted that in this study, biopsy was performed in only five of the cases discussed.

"The results seem most convincing, but would be far strengthened with expert biopsy review of all reported cases in this study," he said.

To date, the dominant theory about the cause of NSF is that transmetallation results in the release of toxic gadolinium from the chelating agent. Linear agents are thought to be more likely to release gadolinium and macrocyclic agents are thought to be more stable (NSF prompts some experts to draw lines between MRI contrast agents).

Kanal and others voiced concerns about the FDA's handling of NSF in a commentary published online in the journal Radiology in mid-September.

Commentary authors included:

• Dr. Dale Broome, head of abdominal radiology, Loma Linda University

• Dr. Diego Martin, professor of radiology, Emory University

• Dr. Henrik Thomsen, professor of radiology, University of Copenhagen

The letter cites evidence showing most NSF cases have been linked to the agent Omniscan and questions FDA policy, which implies equivalent risk for all the agents on the U.S. market.

"The FDA does not acknowledge that the actual risk of developing NSF following Omniscan administration may be higher than that after administration of some or all of the other agents," the authors wrote.

A recently reported case involving administration of the macrocyclic agent ProHance took some observers aback (Bracco reports case of NSF with gadolinium contrast agent ProHance).

The case was acknowledged in writing by Bracco, the agent's manufacturer, in response to an individual's request for information.

Even with a more stable contrast agent, the risk of NSF is not zero, said Alberto Spinazzi, senior vice president of medical and regulatory affairs at Bracco.

"At this juncture, you have to really consider very carefully using gadolinium in high-risk patients. And you must use the lowest dose possible to make a diagnosis," Spinazzi said.

Kanal cautioned that while it was quite likely that NSF may at least theoretically develop after exposure to any gadolinium-based macrocyclic agent, the single ProHance case remains controversial, since it is not clear there was biopsy confirmation or access to detailed medical records involving that patient. Further, per the published reports, there were multiple higher doses of ProHance administered to this patient over a short period of time.

Meanwhile isolated reports of NSF occurring with no exposure to gadolinium are cropping up. In a large, recently published study from Scotland, one patient had no known prior exposure (Radiology 245:1:168-175).

And in the new Mass General study, one NSF patient had no record of receiving a gadolinium-enhanced MRI scan. However, experts say that tracking down exposure to gadolinium is sometimes very difficult, even if scores of hospital records are reviewed. It's possible, for example, that gadolinium might have been used in a CT study or conventional angiography but was not documented.

Today, treatment options for NSF remain very limited. Symptoms are severe and prevention is essential. At the end of August, the ACR published new screening guidelines online to complement MR safety guidelines. Among other things, the guidelines recommend prescreening patients before gadolinium-enhanced MRI and advise a recent glomular filtration rate assessment for at-risk patients.